All regular jet engines and most rocket engines operate on the deflagration of fuel, that is, the rapid but subsonic combustion of fuel. The pulse detonation engine is a concept to create a jet engine that operates on the supersonic detonation of fuel.
The basic operation of the PDE is similar to that of the pulse jet engine; air is mixed with fuel to create a flammable mixture that is then ignited. The resulting combustion greatly increases the pressure of the mixture to approximately 100 atmospheres (10 MPa), which then expands through a nozzle for thrust. To ensure that the mixture exits to the rear, thereby pushing the aircraft forward, a series of shutters are used to close off the front of the engine. Careful tuning of the inlet ensures the shutters close at the right time to force the air to travel in one direction only through the engine.
The main difference between a PDE and a traditional pulse jet engine is that the mixture does not undergo subsonic combustion but instead, supersonic detonation. In the PDE, the oxygen and fuel combination process is supersonic, effectively an explosion instead of burning. The other difference is that the shutters are replaced by more sophisticated valves. In some PDE designs from General Electric, the shutters are eliminated through careful timing, using the pressure differences between the different areas of the engine to ensure the "shot" is ejected rearward.
The main side effect of the change in cycle is that the PDE is considerably more efficient. In the pulse jet engine the combustion pushes a considerable amount of the fuel/air mix (the charge) out the rear of the engine before it has had a chance to burn (thus the trail of flame seen on the V-1 flying bomb). Even while inside the engine the mixture's volume is continually changing, which is an inefficient way to burn fuel. In contrast the PDE deliberately uses a high-speed combustion process that burns all of the charge while it is still inside the engine at a constant volume.
This is said to increase the amount of heat produced per unit of fuel above any other engines, although conversion of that energy into thrust would remain inefficient. A combustion process able to produce more per unit of fuel would, of course, be incredibly valuable in countless applications. Another side effect, not yet demonstrated in practical use, is the cycle time. A traditional pulsejet tops out at about 250 pulses per second due to the cycle time of the mechanical shutters, but the aim of the PDE is thousands of pulses per second, so fast that it is basically continuous from an engineering perspective. This should help smooth out the otherwise highly vibrational pulsejet engine — many small pulses will create less volume than a smaller number of larger pulses for the same net thrust. Unfortunately, detonations are many times louder than deflagrations.
The major difficulty with a pulse-detonation engine is starting the detonation. While it is possible to start a detonation directly with a large spark, the amount of energy input is very large and is not practical for an engine. The typical solution is to use a deflagration-to-detonation transition (DDT)—that is, start a high-energy deflagration, and have it accelerate down a tube to the point where it becomes fast enough to become a detonation. Alternatively the detonation can be sent around a circle and valves ensure that only the highest peak power can leak into exhaust.
This process is far more complicated than it sounds, due to the resistance the advancing wavefront encounters (similar to wave drag). DDTs occur far more readily if there are obstacles in the tube. The most widely used is the "Shchelkin spiral", which is designed to create the most useful eddies with the least resistance to the moving fuel/air/exhaust mixture. The eddies lead to the flame separating into multiple fronts, some of which go backwards and collide with other fronts, and then accelerate into fronts ahead of them.
The behavior is difficult to model and to predict, and research is ongoing. As with conventional pulsejets, there are two main types of designs: valved and valveless. Designs with valves encounter the same difficult-to-resolve wear issues encountered with their pulsejet equivalents. Valveless designs typically rely on abnormalities in the air flow to ensure a one-way flow, and are very hard to achieve in a regular DDT.
NASA maintains a research program on the PDE, which is aimed at high-speed, about Mach 5, civilian transport systems. However most PDE research is military in nature, as the engine could be used to develop a new generation of high-speed, long-range reconnaissance aircraft that would fly high enough to be out of range of any current anti-aircraft defenses, while offering range considerably greater than the SR-71, which required a massive tanker support fleet to use in operation.
While most research is on the high speed regime, newer designs with much higher pulse rates in the hundreds of thousands appear to work well even at subsonic speeds. Whereas traditional engine designs always include tradeoffs that limit them to a "best speed" range, the PDE appears to outperform them at all speeds. Both Pratt & Whitney and General Electric now have active PDE research programs in an attempt to commercialize the designs.
Key difficulties in pulse detonation engines are achieving DDT without requiring a tube long enough to make it impractical and drag-imposing on the aircraft; reducing the noise (often described as sounding like a jackhammer); and damping the severe vibration caused by the operation of the engine.
EDL